The aim of the project is to adapt and to validate the use of a technology developed by Thales to realize for lightning generation to the fragmentation of panels (solar , isolation and radiological). This technology was proved to be efficient on the fragmentation of several kind of materials but no test has been performed on screens/panels, because these requires a full-time team, specialized in the chemistry in aqueous environment wich was not available in Thales at the time of the studies. Within the framework of this program, Thales will use its two prototypes : - One realized, fully sponsored by Thales, operating on the principle of indirect electric arc through water - One realized, during an Eco-industrie 2010 projet, dedicated to the destruction of the small WEE, using the principle of direct arc through water. During this study the scientific teams of Thales, the ENSAM/MAPIE and ENSCP Paristech will try to go deeper into the knowledge of the physical phenomena at stake. These research on the electric, mechanical and chemical effects have for objectives: ·The understanding of the physical phenomena which leads to the fragmentation. The aim is the realization of a better adapted reactor: oModification of electrodes and the shape of wave (study of the electric phenomena) oModification of the shape and the material of the reactor (study of the mechanical phenomena) oChoice of the liquid (chemical study and knowledge of the electric properties of materials) ·The definition of the chain of recycling of the secondary primary materials. oDefinition of the filtering and the research for recycling and for remediation of the solid extracts oDefinition of the methods of recycling the liquid extracts and the separation of the vaporisable materials. This study will allow to assess the gain brought by this technology compared to the means of more classic recycling of panels. There could be several kind of gains: -Economic: oBy recycling and reselling the materials of products at the end of life oBy decreasing the not valued rubbishes -Ecological: oBy decreasing the quantity of buried waste oBy using a concept of "clean" treatment -Design: oAuthorization to use composite panels in an approach of eco design. oSupply new rules of eco-design linked to the new processing procedure and assess the economic and ergonomic gain. Products used for the study are: -Radiologic panels produced by Trixell -lead protected hoods in carbon fiber produced by Trixell -Solar energy panels supplied by the INES -Hight performances Thermic isolation panels supplied by INES The feasibility of the recycling of raw materials will be validated by ENSAM and ENSCP ParisTech. The economic gains will be calculated by each of the suppliers of panels. The ecological gains will be validated by ENSAM and ESCP ParisTech by taking into account treatments being realized on damaged screens and based on studies of global life cycle of produced (gain in the production by recycling rubbishes) The gains in design will be presented by the industrialists on the advices of ENSAM. Then, the industrial specifications of a machine adapted to the problem of panels will be presented by Thales.

Among the large variety of innovative materials engineered for industrial needs, super-hydrophobic surfaces (SHS) have received increasing attention since the nineties. By tuning properly their physicochemical properties, these biomimetic surfaces can entrap a lubricating gas layer within the roughness restricting thereby the solid/liquid contact area. This feature also referred to the “Lotus effect” leads to liquid repellence, which may have a strong impact in engineering applications where wettability control is essential. In the framework of drag reduction, a number of studies have evidenced the beneficial effects of SHS in reducing skin friction at the laboratory scale in well controlled operating conditions. However, predicting their performances under extreme conditions such as highly turbulent flows, which are representative of industrial applications, is a major challenge. In particular, very recent works have pointed out the importance of the deformation of the gas/liquid interface, which is currently neglected in the design of SHS. The ambition of the IDEFHYX project is precisely to overcome this lack by addressing the problem with a radically different approach from those used until now. The originality of this project is based on the coupling between two-phase flow simulations, stability analysis and model experiments to provide a more in-depth understanding of the physical mechanisms governing the interaction of the gas/liquid interface and its surrounding environment. Doing so, IDEFHYX will provide a unique framework on which the next generation of SHS will be engineered coupling effectiveness and resilience even in extreme conditions.

Data-driven modeling has been having a transformative impact in many scientific fields. Of particular interest here are control-oriented techniques such as POD, ERA, or DMD. Despite their widespread usage, such techniques cannot easily provide a faithful representation of the system under scrutiny when the latter is parameterized. The aim of the present project is to tackle this issue. By reformulating these different techniques in the framework of reduced-rank regression, an optimal data-driven SVD factorization of the underlying operator can be obtained. Building on recent works about interpolation on matrix manifold, this factorization can be interpolated based on the Stiefel manifold, i.e. the manifold of orthogonal matrices. Using these new mathematical tools, the aim of the present project is to explore our ability to obtain parameterized data-driven models with a particular emphasis on closed-loop feedback control in fluid dynamics and to quantify the accuracy and robustness of these models. In order to illustrate our methodology, two classical problems from flow control (cylinder and shear-driven cavity flow) will be considered. In both cases, different flavours of reduced-order models will be considered (e.g. ERA, DMD, BPOD, etc) and parametric dependencies built into them. Because using the Stiefel manifold equips us with a natural notion of distance between two modal decompositions, efficient sampling strategies of the parameter space can be devised. Given data-driven models obtained at different points in the parameter space, an approximate model can be obtained for a new query point with minimal computational complexity using such interpolation schemes. Given the ubiquity of matrix decompositions in engineering sciences, the methodology proposed in this project is expected to have far-reaching applications, not only in fluid dynamics but also in data science and machine learning in general.

The project addresses the interior noise generated by the turbulent boundary layer around modern transportation vehicle such as aircrafts, high speed trains and automotives. The excitation by the wall turbulence usually constitutes a major internal noise contribution during cruising trip. The random character of the pressure distribution induced by the boundary layer has a major effect on the noise transmitted through the fuselage and its cross-spectral density plays a major role in determining the effective force causing motion to the structure. The objectives of the project are twofold: from a more fundamental point of view, the question of the contribution of the acoustic part of the pressure fluctuations will be raised. In particular, the comparison between a well-controlled experimental database and simulations (using Direct Noise Computations) should help to clarify some issues. One of the most important goals is to provide a non-ambiguous quantification of pressure levels in the acoustic domain. The role of shear stresses and the behaviour at very low wave numbers are also fundamental issues considered in the present project. Hypotheses made in semi-empirical modelling will be investigated and improved with regard to experimental and numerical databases. From a more industrial point of view, the reliability of existing semi-empirical models used for design is uncertain, notably for high subsonic Mach numbers characteristic of aeronautic applications (M=0.8), and for non-equilibrium boundary layers, where the spatial homogeneity assumption is broken by a pressure gradient or a surface discontinuity. Simulation tools on academic cases can be used to help understanding the modifications of the excitation sources. A validation step is crucial where experiments and DNC are compared in a similar flow configuration. DNC simulations can then be used to tackle M=0.8 configuration, the effect of a pressure gradient, and the presence of a forward or backward facing step. The analysis of the results and improvement of semi-empirical model should provide the designer with predictive tools for the boundary layer induced noise over representative structures and realistic Mach conditions. The project will thus deliver enhanced models (numerical and semi-empirical) for the pressure fluctuations beneath a turbulent boundary layer (notably in the acoustic domain and for the effect of pressure gradients) over complex structures. Guidelines for the industrial use will be derived.